Loading...

Feasibility Study of Using the Elemental Analysis of Prompt Gamma Spectrum to Improve the Treatment Planning in Hadron Therapy

Saheli, Fereshte | 2021

341 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 54272 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Vosoughi, Naser; Riazi, Zafar; Rasouli, Fateme
  7. Abstract:
  8. Hadron therapy is one of the cancer treatment methods using the targeted dose distributhion. In hadron therapy, the prompt gamma is produced from excited nucleas of target in the following of non-elastic nuclear interactions between the target and the incident proton within few nano-seconds and with energy less than 10 MeV. The excited energy level depends on incident particle energy and the target materials. Since the Spatial distribution of prompt gamma rays depends on incident particles energy, it can be useful for determinding the incident particle range. Also, the prompt gamma energy spectrum of each element is an individual feature, thereby targets with different composition of elements produce dedicated prompt gamma spectra during hadron therapy. In this thesis, elemental analysis of irradiated tissue in proton therapy is done using two different approaches. One of them is single peak analysis and the other one is whole spectrum analysis. In whole spectrum analysis method, it is assumed that the detector spectrum is a linear combination of contributing fundamental library spectra of elemental composition in an unknown sample. This approach is implemented using two optimization methods: the multiple linear regression (MLR) method and a genetic algorithm (GA). The Geant4.10.03 toolkit is applied to simulate single element and multi-elements phantoms for evaluating the feasibility of PG spectroscopy in accurate elemental analysis of tissues. In single peak analysis method, the least square and neural network methods are used to estimate weight percent of elements. The results of whole spectrum analysis demonstrated that the majer challenge of this approach is to simulate accurate library spectra in each irradiated phantom. For example, quantitative results of determining oxygen concentration, as a critical element in the human body, using both GA and MLR methods are estimated with an error less than 1.5% for 30, 70 and 150 MeV proton energy in the all considered phantoms. The estimated error decreased by exploiting the detector shield and collimator in this work. The single peak analysis approach produces suitable results in both linear mapping and neural network methods as well. In linear mapping method, the estimated error of oxygen consentration is respectively less than 3.5 and 7% for 30 and 120 MeV proton energy in all considered phantoms. In neural network method, the corresponding error of oxygen element is estimated less than 5.8% for tumors considered in the human eye phantoms. Moreover, the estimated error of other main elements of human body is reported in section 6. In conclution, the potential of performing elemental analysis using prompt gamma ray spectrum of irradiated phantom and the correlation between the obtained elemental analysis results with incident particle range are confirmed in this thesis
  9. Keywords:
  10. Proton Therapy ; Monte Carlo Model ; Element Analysis ; Gamma Ray Spectrometry ; Whole Spectrum Analysis ; Single Peak Analysis

 Digital Object List

 Bookmark

No TOC